Tensors and Dynamic neural networks in Python with strong GPU acceleration
-
Updated
Jun 1, 2024 - Python
Artificial neural networks (ANN) are computational systems that "learn" to perform tasks by considering examples, generally without being programmed with any task-specific rules.
Tensors and Dynamic neural networks in Python with strong GPU acceleration
An Open Source Machine Learning Framework for Everyone
🦋 A personal research and development (R&D) lab that facilitates the sharing of knowledge.
NNtrainer is Software Framework for Training Neural Network Models on Devices.
On-device AI across mobile, embedded and edge for PyTorch
A flexible, high-performance serving system for machine learning models
A tool for cell instance aware segmentation in densely packed 3D volumetric images
Profile PyTorch models for FLOPs and parameters, helping to evaluate computational efficiency and memory usage.
Visualizer for neural network, deep learning and machine learning models
Implementation for the different ML tasks on Kaggle platform with GPUs.
An ASL detection script utilizing a TensorFlow image classification model trained from scratch. It is tailored to recognize American Sign Language (ASL) alphabet letters from live video streams, and provides documentation covering the neural network architecture, installation, dataset details, training procedures, and real-time detection.
High-efficiency floating-point neural network inference operators for mobile, server, and Web
Burn is a new comprehensive dynamic Deep Learning Framework built using Rust with extreme flexibility, compute efficiency and portability as its primary goals.
Resources needed to start deep learning research. ML/DL/CV/NLP/ML-SYS/RL/Graphs/Maths/Med Img lecture videos from professors at esteemed universities.
A strong chess engine made in c++
This repository offers a straightforward implementation of Vision Transformers (ViT), specifically designed for computer vision tasks using PyTorch. Dive into efficient and practical transformer applications for image recognition.
Tensor parallelism is all you need. Run LLMs on weak devices or make powerful devices even more powerful by distributing the workload and dividing the RAM usage.
This project is an introduction to artificial neural networks thanks to the implementation of a multilayer perceptron.
Qdrant - High-performance, massive-scale Vector Database for the next generation of AI. Also available in the cloud https://cloud.qdrant.io/